• Health Policy and Government Relations
• Contact Congress - Take Legislative Action
• SVS Health Policy Committee Meets with Senators/Staff
• Priority Issues
• Imaging Cuts
• SAAAVE Act
• Physician Payment
• Pay For Performance (P4P)
• Fistula First Initiative
• Malpractice Reform
• Reimbursement
• Regulatory Issues
• Political Action Committee
• 110th Congress Committee Assignments

P4P: Measuring Outcomes

Perhaps the most important steps in any P4P program are data collection, aggregation, adjustment, analysis, and reporting. Interestingly enough, however, these are the components that have not been well developed yet. Data collection is primitive, relying on voluntary programs and G-codes. Data aggregation is very limited in its scope. Except for very few examples, data registries that can be used on the national level are almost nonexistent. Available registries are not uniform, and each collects different data points. Programs that use risk adjustment are not available except for the VA and, now, the ACS National Surgical Quality Improvement Program (NSQIP) and the Society for Thoracic Surgery’s Cardiac Surgery Program.

National Surgical Quality Improvement Program

There is no question that, for many years, the VA has been on the forefront in establishing the premier national quality-improvement program in the nation. This program was initiated in response to a 1986 congressional law that mandated the VA to report its surgical results annually comparing risk-adjusted outcomes with the national average. Interestingly, Congress overlooked the fact at the time that there were no risk-adjusted national outcomes to compare to. This has not changed, and NSQIP remains the first and only noncardiac, national, risk-adjusted, validated, peer-controlled, and outcome- based system; it has been in existence for over 15 years.

NSQIP has proven to enhance positively the quality of care of surgical patients in the VA system. From the inception of this program in 1991 to 2002, the 30-day incidence of mortality decreased by 27% and morbidity by 45% in participating VA medical centers.¹⁸ Data on every surgical procedure in VA Medical Centers performed under general, epidural, or spinal anesthesia are collected; in addition, data on specific operations are included regardless of the type of anesthesia used, such as carotid endarterectomy (CEA) performed under local or regional anesthesia. Fifty-two preoperative variables (10 demographic, 30 clinical, and 12 laboratory variables), 15 clinical intraoperative variables, and thirty-three 30-day outcome variables (to include 30-day mortality and hospital length of stay) are collected by a nurse clinical reviewer from electronic medical records on about 115,000 of the total procedures performed in 128 VA hospitals annually. The data are adjusted on the basis of the preoperative risk factors transmitted to the Data Coordinating Center. The rationale for data adjustment and analysis is based on a conceptual framework in which outcomes of health care are determined by three factors: namely, quality of care, patient risk factors, and random variation. If patient’s risk factors are accounted for by risk adjustment and random variation is accounted for by statistical analysis, then outcome accounts for quality of care. The risk-adjusted mortality and morbidity for the particular group of patients are reported as expected. The ratios of observed over expected mortality and morbidity (O/E ratio) are reported to chiefs of surgery on a quarterly basis through a brief report. Annually, a voluminous report is sent to the Chief of Surgery and the leadership of each medical center; it includes data collection, analysis, and, most importantly, comparison with other VA medical centers around the country. The practices of low-outlier medical centers with O/E ratio significantly below 1.0 are considered best practices and are fed back to other medical centers for their consideration and implementation. The validation study has shown that high outliers have inferior structures and processes of care when compared with low outliers.^18-20 The NSQIP Executive Committee reviews on an annual basis the preceding 4 years’ cumulative results; those medical centers that are consistently high outliers are asked to review specific areas of concern or are visited by a team that would analyze their programs and offer suggestions for change.

NSQIP’s leadership is considering future directions for the program, which may include collecting data on longterm survival, functional outcomes, quality of life, patients’ satisfaction, and cost-effectiveness, to name a few. In addition, it is considering the development of process measures that directly affect outcome.¹⁸

Recognizing the importance and the impact of NSQIP on veterans’ health care, private institutions expressed interest in extending this database to nonfederal hospitals. This prompted a pilot study in 1999 to alpha test the program in three non-VA hospitals (University of Michigan, Emory University, and University of Nebraska). When the applicability of this program was proven in this pilot study, the ACS, through a grant from AHRQ, extended this program to 11 additional hospitals in a beta test study. This program, ACS-NSQIP, has been implemented in over 100 hospitals nationwide. In addition, procedure-specific modules based on the NSQIP database, such as bariatric surgery, are being developed by the College in collaboration with specialty societies. Already, NSQIP is serving as a P4P platform for private sector initiatives. Blue Cross/Blue Shield of Michigan started in November of 2005 paying 15 hospitals to participate in ACS-NSQIP and contributing
data on general and vascular procedures. This project was an extension of another one that started in 1997 in which Blue Cross/Blue Shield of Michigan paid part of the cost to Michigan hospitals for providing data on percutaneous coronary interventions. During this pay for participation program, no site- or provider-specific data were revealed to Blue Cross/Blue Shield of Michigan or the general public; the data were used by institutions to create quality-improvement measures. This project resulted in a significant drop in complications and a 27% drop in mortality from coronary interventions.²¹ ACS-NSQIP seems to be the most logical platform to be used for CMS’ SCIP program. CMS and its former administrator, Dr McLellan, are very interested in NSQIP migration to the private sector. It is estimated that NSQIP will become the platform for CMS P4P initiatives.²²

SVS Carotid Stenting and Endarterectomy Registry

In direct response to CMS’ reimbursement policy for carotid stenting, which required hospitals to maintain ongoing data and information about these procedures, the Outcomes Committee of the SVS, headed by past President Dr. Greg Sicard and supported by the leadership of the SVS, urgently put together the SVS carotid stenting and CEA registry. This database is designed to comply with CMS payment requirements and collect long-term data on carotid stenting and CEA. Currently, this is the only registry that incorporates CEA data in addition to stenting. The registry is Health Insurance Portability and Accountability Act compliant and Web based and provides the ability of designated 7 individuals in each of the participating medical centers to obtain real-time reports of their center’s results with comparison to national averages and other centers in a blinded fashion. Institutions that participate in this database are ready to provide CMS with the data required for their recredentialing.²³

Article References

Reprinted from Journal of Vascular Surgery, Volume 44, Number 4, October 2006, Anton M. Sidawy, MD, MPH, Washington, D.C., "Pay for Performance: The process and its evolution," 892-902, Copyright 2006, with permission from The Society for Vascular Surgery.

Journal of Vascular Surgery (http://www.jvascsurg.org)